The Art of Reading Prediction Markets: The Information Factor

by PredictionNews Editorial Staff

May 13, 2025

Trade on Prediction Markets » The Art of Reading Prediction Markets: The Information Factor

Reading prediction market prices is an art, where deciding which prices to take seriously requires understanding of many nuances. Some markets are based on a topic too niche to pull useful information together. Others have low trading volume but still produce a useful price.

Rutgers Statistics Professor Harry Crane studies prediction markets and built a model that did a better job of predicting the 2024 election than many polls. Crane spoke with Prediction News to explain the differences that set useful prediction markets apart from those that have little meaning.

Lorem Ipsum is simply dummy text of the printing and typesetting industry. Lorem Ipsum has been the industry’s standard dummy text ever since the 1500s, when an unknown printer took a galley of type and scrambled it to make a type specimen book.

Latest Political Prediction News

News

DOGE Report Card: Markets Tell The Real Story as Elon Musk Exits

by Chris Gerlacher

June 5, 2025

0

Elon Musk officially left government service last Friday, receiving his medal after attempting to cut $2 trillion in government spending....

Lorem Ipsum is simply dummy text of the printing and typesetting industry. Lorem Ipsum has been the industry’s standard dummy text ever since the 1500s, when an unknown printer took a galley of type and scrambled it to make a type specimen book.

It has survived not only five centuries, but also the leap into electronic typesetting, remaining essentially unchanged. It was popularised in the 1960s with the release of Letraset sheets containing Lorem Ipsum passages, and more recently with desktop publishing software like Aldus PageMaker including versions of Lorem Ipsum.

Trending Political Betting Markets

How To Bet On Politics

Even a prediction market price that comes from a wise crowd isn’t necessarily a source of “truth.” Crane offers a more nuanced interpretation using a 50-cent price as an example:

“It [a prediction market price] means that the best information that’s [contained in] the aggregated information of all the participants in the market suggests that 50/50 is the best price.”

Some prediction markets are more useful than others

Prediction markets work best when there is a large amount of information to bring together. The vast amount of available information made election markets so effective at identifying Donald Trump’s edge in the race against Kamala Harris.

Crane offered Polymarket’s market on the contents of Luigi Mangione’s manifesto would be released as a counterexample. Even though the market’s total trade volume reached almost $280,000, some of the markets on specific terms attracted less than $10,000, a minuscule amount compared to Polymarket’s $3.7 billion U.S. election market.

“In a market like that [the Mangione manifesto market], there’s a small amount of people interested in it,” Crane said. “There’s also not very much information to go on, right? I mean, you either know what’s in it or you don’t know what’s in it, and so there’s really not a whole lot of information that [the] market can aggregate.”

Even if each of the mention markets had arrived at a “true” price, the odds wouldn’t be very informative. Mangione’s mention of “Biden” or “woke” wouldn’t have said much about what was likely to be an illogical piece of writing. (All of the choices in that market resolved to No, anyway.)

“That’s a case where you know that market is not all that useful, not nearly as good at aggregating information as these election markets are, which I think are quite, quite good,” Crane said.

How prediction markets beat the polls

Unlike the Luigi Mangione market, election markets had high trade volume, high interest, and a vast amount of public information to pull into one price. Crane built an election model that used multiple prediction markets to forecast Trump’s chance of winning at 67%, ahead of individual prediction markets and ahead of polls that suggested a 50/50 race.

Political polls have struggled to accurately measure Trump’s support during his presidential career. During the last three presidential elections, political polls underestimated Trump’s support in many states. Polls incorrectly predicted Hillary Clinton’s victory in 2016, overestimated Joe Biden’s margin of victory in 2020, and underestimated Trump’s support in the 2024 swing states.

Crane attributes these errors to the “demographic corrections” polls must make to measure the electorate.

“A lot of their corrections are based on what happened last election, and that does work pretty [well]. It gets you pretty close. But where it matters is kind of in this margin of ‘What is the difference between this election and last election?’ That’s always the question that we want, and that’s where it starts to get a bit harder to answer.”

Pollsters have to weigh different groups of people to make sure their samples match the electorate. The need to match is why incorrect weighting can paint a wildly different picture of the election.

Prediction markets strengths and weaknesses in election forecasting

In contrast, prediction markets can aggregate voters’ and analysts’ opinions by letting every trader bring their own information to the market. Prediction markets only have to attract a large number of people who can cancel each other’s errors instead of hoping the “correct” number of Republicans participate and publicly admit their support for Trump.

Bernard College Economics Professor Rajiv Sethi compared Polymarket, The Economist, and FiveThirtyEight forecasts leading up to the 2024 election in a Substack post. He found that FiveThirtyEight had the best forecasts of the House races he analyzed, but Polymarket outperformed the Economist.

“The ability of markets to absorb novel sources of information and a broad range of perspectives is both a strength and a weakness—the market mechanism can perform much better than models when history is a poor guide to the future, but it can also exhibit overreaction and excess volatility,” Sethi wrote.

Sethi cites the market shock from Ann Selzer’s poll showing Harris ahead by 6 in Iowa. Trump’s odds of winning fell more in Polymarket’s contracts on Iowa’s first and third districts while FiveThirtyEight and The Economist’s models were “much more modest.”

Some prices are worth more than others

When reading a prediction market price, it’s important to remember that the price is the traders’ best interpretation of the publicly available information. Even if a few traders are influenced by information from their own personal bubble, those insights have been priced into their trades and should be available to anyone in their position. In a market with important non-public information, like mention markets on Trump’s speeches, odds can be particularly volatile and limited in their forecasting ability.

The market itself matters too. If the topic is too niche, then the price may not mean very much. Trader interest and public information quality are crucial to identifying prediction market prices that can be taken seriously and those that must be taken with not just a grain of salt but a recognition that the market may be barely keeping up instead of accurately looking ahead.

Prediction Market Websites

Kalshi

Polymarket

PredictIt

Manifold

ForecastEx

Metaculus

Facebook X Linkedin

PredictionNews Editorial Staff

Articles published by PredictionNews editorial staff reflect contributions from three or more writers. We explore prediction industry trends, keeping you informed about prediction market data, price movements, and information that inform trading strategies. This rapidly evolving industry produces news 24x7, and our editorial team distills it for you.